trouble migrating LVM from IDE to SATA
philsf at ufrj.br
Sat Nov 4 19:11:35 UTC 2006
I am trying to create two VGs in my new SATA2 HD to replace my IDE HD, but
vgcreate refuses to create anything, as it complains that the PVs are not
PVs. That is in spite of pvcreate succesfully reporting it created the PVs. I
tried several times creating and removing. Follow results from -d (debug)
and -v (verbose) of creation process.
philsf at peralta:~$ sudo pvcreate -v -d /dev/sda
Set up physical volume for "/dev/sda5" with 39069633 available sectors
Zeroing start of device /dev/sda5
Physical volume "/dev/sda5" successfully created
Set up physical volume for "/dev/sda6" with 39069633 available sectors
Zeroing start of device /dev/sda6
Physical volume "/dev/sda6" successfully created
Set up physical volume for "/dev/sda7" with 271353468 available sectors
Zeroing start of device /dev/sda7
Physical volume "/dev/sda7" successfully created
After which I tried:
philsf at peralta:~$ sudo vgcreate vg /dev/sda5 /dev/sda6 /dev/sda7
No physical volume label read from /dev/sda5
/dev/sda5 not identified as an existing physical volume
Unable to add physical volume '/dev/sda5' to volume group 'vg'.
Now when I see this, I can't determine if the problem is with vgcreate not
recognizing the PVs, or pvcreate not creating them correctly. So I tried the
philsf at peralta:~$ sudo pvscan
and the only PVs that appear, are the ones in use (from my IDE hard drive). I
can't seem to find any option in pvscan to force it to look in /dev/sda, but
so far, it really seems the partitions aren't PVs yet, right? Wrong. If out
of curiosity I issue another "create", I get an error:
philsf at peralta:~$ sudo pvcreate /dev/sda7
Failed to write physical volume "/dev/sda7"
And I am able to pvremove it normally (only once). So my guess is that the
PV's are indeed getting created, but not recognized later. This only happens
in my SATA hard drive.
I am kind of lost here, anybody knows what's going on?
More information about the ubuntu-users